perf: backport CountGroupsAccumulator

Author: srh

datafusion/src/physical_plan/expressions/count.rs

@@ -20,13 +20,16 @@
 use std::any::Any;
 use std::sync::Arc;
 
-use crate::error::Result;
-use crate::physical_plan::groups_accumulator::GroupsAccumulator;
+use crate::error::{DataFusionError, Result};
+use crate::physical_plan::groups_accumulator::{EmitTo, GroupsAccumulator};
 use crate::physical_plan::groups_accumulator_flat_adapter::GroupsAccumulatorFlatAdapter;
+use crate::physical_plan::null_state::accumulate_indices;
 use crate::physical_plan::{Accumulator, AggregateExpr, PhysicalExpr};
 use crate::scalar::ScalarValue;
+use arrow::array::{Array, ArrayData, BooleanArray, Int64Array, PrimitiveArray};
+use arrow::buffer::Buffer;
 use arrow::compute;
-use arrow::datatypes::DataType;
+use arrow::datatypes::{DataType, Int64Type, UInt64Type};
 use arrow::{
     array::{ArrayRef, UInt64Array},
     datatypes::Field,
@@ -99,11 +102,7 @@ impl AggregateExpr for Count {
     fn create_groups_accumulator(
         &self,
     ) -> arrow::error::Result<Option<Box<dyn GroupsAccumulator>>> {
-        Ok(Some(Box::new(GroupsAccumulatorFlatAdapter::<
-            CountAccumulator,
-        >::new(move || {
-            Ok(CountAccumulator::new())
-        }))))
+        Ok(Some(Box::new(CountGroupsAccumulator::new())))
     }
 
     fn name(&self) -> &str {
@@ -170,6 +169,201 @@ impl Accumulator for CountAccumulator {
     }
 }
 
+/// An accumulator to compute the counts of [`PrimitiveArray<T>`].
+/// Stores values as native types, and does overflow checking
+///
+/// Unlike most other accumulators, COUNT never produces NULLs. If no
+/// non-null values are seen in any group the output is 0. Thus, this
+/// accumulator has no additional null or seen filter tracking.
+#[derive(Debug)]
+struct CountGroupsAccumulator {
+    /// Count per group.
+    ///
+    /// Note that in upstream this is a Vec<i64>, and the count output and intermediate data type is
+    /// `DataType::Int64`.  But here we are still using UInt64.
+    counts: Vec<u64>,
+}
+
+impl CountGroupsAccumulator {
+    pub fn new() -> Self {
+        Self { counts: vec![] }
+    }
+}
+
+impl GroupsAccumulator for CountGroupsAccumulator {
+    fn update_batch(
+        &mut self,
+        values: &[ArrayRef],
+        group_indices: &[usize],
+        opt_filter: Option<&BooleanArray>,
+        total_num_groups: usize,
+    ) -> Result<()> {
+        assert_eq!(values.len(), 1, "single argument to update_batch");
+        let values = &values[0];
+
+        // Add one to each group's counter for each non null, non
+        // filtered value
+        self.counts.resize(total_num_groups, 0);
+        accumulate_indices(
+            group_indices,
+            values.data_ref().null_bitmap().as_ref().map(|bitmap| (bitmap, values.offset(), values.len())),
+            opt_filter,
+            |group_index| {
+                self.counts[group_index] += 1;
+            },
+        );
+
+        Ok(())
+    }
+
+    fn merge_batch(
+        &mut self,
+        values: &[ArrayRef],
+        group_indices: &[usize],
+        opt_filter: Option<&BooleanArray>,
+        total_num_groups: usize,
+    ) -> Result<()> {
+        assert_eq!(values.len(), 1, "one argument to merge_batch");
+        // first batch is counts, second is partial sums
+        let partial_counts = match values[0].as_any().downcast_ref::<PrimitiveArray<UInt64Type>>() {
+            Some(x) => x,
+            None => {
+                panic!("values[0] is of unexpected type {:?}, expecting UInt64Type for intermediate count batch", values[0].data_type());
+            }
+        };
+
+        // intermediate counts are always created as non null
+        assert_eq!(partial_counts.null_count(), 0);
+        let partial_counts = partial_counts.values();
+
+        // Adds the counts with the partial counts
+        self.counts.resize(total_num_groups, 0);
+        match opt_filter {
+            Some(filter) => filter
+                .iter()
+                .zip(group_indices.iter())
+                .zip(partial_counts.iter())
+                .for_each(|((filter_value, &group_index), partial_count)| {
+                    if let Some(true) = filter_value {
+                        self.counts[group_index] += partial_count;
+                    }
+                }),
+            None => group_indices.iter().zip(partial_counts.iter()).for_each(
+                |(&group_index, partial_count)| {
+                    self.counts[group_index] += partial_count;
+                },
+            ),
+        }
+
+        Ok(())
+    }
+
+    fn evaluate(&mut self, emit_to: EmitTo) -> Result<ArrayRef> {
+        let counts = emit_to.take_needed(&mut self.counts);
+
+        // Count is always non null (null inputs just don't contribute to the overall values)
+
+        // TODO: This copies.  Ideally, don't.  Note: Avoiding this memcpy had minimal effect in PrimitiveGroupsAccumulator
+        let buffers = vec![Buffer::from_slice_ref(&counts)];
+
+        let data = ArrayData::new(
+            DataType::UInt64,
+            counts.len(),
+            None,
+            None,
+            0, /* offset */
+            buffers,
+            vec![]
+        );
+        Ok(Arc::new(PrimitiveArray::<UInt64Type>::from(data)))
+    }
+
+    // return arrays for counts
+    fn state(&mut self, emit_to: EmitTo) -> Result<Vec<ArrayRef>> {
+        let counts = emit_to.take_needed(&mut self.counts);
+        // Backporting note:  UInt64Array::from actually does copy here in old DF.
+        let counts: PrimitiveArray<UInt64Type> = UInt64Array::from(counts); // zero copy, no nulls
+        Ok(vec![Arc::new(counts) as ArrayRef])
+    }
+
+    /// Converts an input batch directly to a state batch
+    ///
+    /// The state of `COUNT` is always a single Int64Array (backporting note: it is a UInt64Array):
+    /// * `1` (for non-null, non filtered values)
+    /// * `0` (for null values)
+    fn convert_to_state(
+        &self,
+        _values: &[ArrayRef],
+        _opt_filter: Option<&BooleanArray>,
+    ) -> Result<Vec<ArrayRef>> {
+        // convert_to_state only gets used in upstream datafusion, and we set
+        // supports_convert_to_state to false. Because values.data_ref().offset() and the null
+        // bitmap have differences that require care to backport, we comment this out instead.
+        return Err(DataFusionError::NotImplemented("Input batch conversion to state not implemented".to_owned()));
+        /*
+        let values = &values[0];
+
+        let state_array = match (values.logical_nulls(), opt_filter) {
+            (None, None) => {
+                // In case there is no nulls in input and no filter, returning array of 1
+                Arc::new(Int64Array::from_value(1, values.len()))
+            }
+            (Some(nulls), None) => {
+                // If there are any nulls in input values -- casting `nulls` (true for values, false for nulls)
+                // of input array to Int64
+                let nulls = BooleanArray::new(nulls.into_inner(), None);
+                compute::cast(&nulls, &DataType::Int64)?
+            }
+            (None, Some(filter)) => {
+                // If there is only filter
+                // - applying filter null mask to filter values by bitand filter values and nulls buffers
+                //   (using buffers guarantees absence of nulls in result)
+                // - casting result of bitand to Int64 array
+                let (filter_values, filter_nulls) = filter.clone().into_parts();
+
+                let state_buf = match filter_nulls {
+                    Some(filter_nulls) => &filter_values & filter_nulls.inner(),
+                    None => filter_values,
+                };
+
+                let boolean_state = BooleanArray::new(state_buf, None);
+                compute::cast(&boolean_state, &DataType::Int64)?
+            }
+            (Some(nulls), Some(filter)) => {
+                // For both input nulls and filter
+                // - applying filter null mask to filter values by bitand filter values and nulls buffers
+                //   (using buffers guarantees absence of nulls in result)
+                // - applying values null mask to filter buffer by another bitand on filter result and
+                //   nulls from input values
+                // - casting result to Int64 array
+                let (filter_values, filter_nulls) = filter.clone().into_parts();
+
+                let filter_buf = match filter_nulls {
+                    Some(filter_nulls) => &filter_values & filter_nulls.inner(),
+                    None => filter_values,
+                };
+                let state_buf = &filter_buf & nulls.inner();
+
+                let boolean_state = BooleanArray::new(state_buf, None);
+                compute::cast(&boolean_state, &DataType::Int64)?
+            }
+        };
+
+        Ok(vec![state_array])
+        */
+    }
+
+    fn supports_convert_to_state(&self) -> bool {
+        // Is set to true in upstream (as it's implemented above in upstream).  But convert_to_state
+        // is not used in this branch anyway.
+        false
+    }
+
+    fn size(&self) -> usize {
+        self.counts.capacity() * size_of::<usize>()
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;